jupes/agent-forge-harness / .agents/skills/go
go
VerifiedClassify a task and route it to the right workflow (fix / feature / epic). Use when the user runs /go or wants the adaptive workflow router.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 4.1 KB |
| agents/openai.yaml | config | 213 B |
SKILL.md
View on GitHub---
name: go
description: Classify a task and route it to the right workflow (fix / feature / epic). Use when the user runs /go or wants the adaptive workflow router.
---
# /go — Adaptive Workflow Router
Single entry point for all development work. Classifies scope and routes to the right workflow.
## Usage
```
/go # Pick highest-priority ready task
/go T-42 # Work on specific task
/go EPIC-7 # Run epic workflow
/go fix login redirect bug # Describe work in free text
/go path/to/spec.md # Read spec file, materialize Beads tasks
/go --epic <description> # Force epic workflow
```
---
## Input Handling
### Empty input
```bash
bd ready
```
Pick the first unblocked, unclaimed task. If none: report "No ready work. Check `bd ready` or create new tasks."
### Task/Epic ID (e.g., `T-42`, `EPIC-7`)
```bash
bd show <id>
```
Fetch details, then classify scope and route to workflow.
### File path
Read the file as a spec. Materialize Beads tasks from it (see Spec Materialization below), then route.
### Free text
Use the description directly. Classify scope, create a Beads task if needed, then route.
### `--epic` flag
Force epic workflow regardless of scope classification.
---
## Scope Classification
Evaluate the task to determine which workflow to use:
| Tier | Criteria | Routes To |
|------|----------|-----------|
| **Fix** | Single task, ≤3 files touched, clear scope, no new architecture | `workflows/fix.md` |
| **Feature** | Single task, >3 files OR needs a plan OR new system/component | `workflows/feature.md` |
| **Epic** | Epic ID, `--epic` flag, or multiple tasks from spec | `workflows/epic.md` |
**When in doubt about Fix vs Feature**: check whether a plan would help. If yes → Feature.
---
## Spec Materialization
When input is a file path, parse the spec and create Beads tasks:
1. Read the file
2. Identify distinct work items (look for headers, numbered lists, bullet points with verbs)
3. For each item:
```bash
bd create \
--type <bug|feature|task|chore> \
--title "[<repo from spec or ask>] <derived title>" \
--priority <value> \
--acceptance "<acceptance criterion if identifiable>"
```
Set **`<value>`** using `.claude/skills/beads-priority-assignment/SKILL.md`.
4. If items have clear dependencies (item B requires item A), add them:
```bash
bd dep add <B-id> <A-id> # B depends on A: the later item first, the earlier item second
```
5. If there are 3+ related items: create an epic to group them
```bash
bd create --type epic --title "<overall goal>" --priority <value>
```
Epic **`<value>`**: same skill — reflect overall urgency of the group.
6. Report what was created, then route
---
## PR Stack Detection
Before routing, check for PR stacks (existing open PRs for the repo):
```bash
gh pr list --state open --json number,title,baseRefName,headRefName
```
If the current branch is part of a stack (a PR whose base is another feature branch):
- Note the stack structure
- When shipping later, use the correct base branch for the PR, not `main`
---
## Routing
After classification, invoke the appropriate workflow:
```
→ Fix: Follow .claude/workflows/fix.md
→ Feature: Follow .claude/workflows/feature.md
→ Epic: Follow .claude/workflows/epic.md
(Adopt the Lead Agent role from .claude/agents/lead.md)
```
Announce routing decision before starting:
```
Routing to: [Fix/Feature/Epic] workflow
Task: <id> — <title>
Reason: <1 sentence why this tier>
```
---
## Error Handling
- **Task not found**: `bd show <id>` returns empty → check ID, suggest `bd ready`
- **Ambiguous spec**: >10 items from a single spec → ask user to confirm before materializing
- **No git repo**: abort with instructions to `git init`
- **Beads unreachable** (`bd` errors / Dolt server down): **STOP — do not proceed without Beads.** Run `bd dolt start` and retry; Beads tracking is mandatory for tracked work. (The SessionStart hook normally auto-starts the server.) Only if `bd` is genuinely not installed should you tell the user to install it before continuing.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| planning-and-task-breakdownaddyosmani/agent-skillsBreaks work into ordered tasks. Use when you have a spec or clear requirements and need to break work into implementable tasks. Use when a task feels too large to start, when you need to estimate scope, or when parallel work is possible. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| screenpipethedotmack/claude-memSearch screen and audio history recorded by Screenpipe and optionally save selected context into claude-mem. Use when the user asks to recall desktop activity, a meeting, or something seen in another app through Screenpipe. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| contact-widgetnexu-io/open-designSelf-contained floating chat widget with welcome screen, social links, meeting button, and message input. Single HTML file, zero dependencies. | Stars: 108k | Updated | Claude CodeCodex | Verified | Install |
| meeting-notesnexu-io/open-designMeeting notes page — title bar with attendees, agenda checklist, decisions block, action items table with owners + dates, and a "next meeting" footer. Use when the brief mentions "meeting notes", "minutes", "1:1 notes", "all-hands recap", or "会议纪要". | Stars: 108k | Updated | Claude CodeCodex | Verified | Install |